US20070114632A1 - Transistor having dielectric stressor elements at different depths from a semiconductor surface for applying shear stress - Google Patents

Abstract

A chip is provided which includes an active semiconductor region and a field effect transistor (“FET”) having a channel region, a source region and a drain region all disposed within the active semiconductor region. The FET has a longitudinal direction in a direction of a length of the channel region, and a transverse direction in a direction of a width of the channel region. A buried dielectric stressor element has a horizontally extending upper surface at a first depth below a major surface of a portion of the active semiconductor region, such as an east portion of the active semiconductor region. A surface dielectric stressor element is disposed laterally adjacent to the active semiconductor region at the major surface of the active semiconductor region. The surface dielectric stressor element extends from the major surface to a second depth not substantially greater than the first depth. The stresses applied by the buried and surface dielectric stressor elements cooperate together to apply a shear stress to the channel region of the FET.

Description

BACKGROUND OF THE INVENTION
• The present invention relates to semiconductor devices and processing. More particularly, the invention relates to semiconductor devices with dielectric stressor elements and methods of making the same.
• A compressive stress or tensile stress can be applied to some types of transistors to increase their performance. In particular, the performance of a p-type field effect transistor (“PFET”) improves when a longitudinal (in the direction of current flow) compressive stress is applied to the channel region. On the other hand, the performance of an n-type field effect transistor (“NFET”) improves when a longitudinal tensile stress is applied to the channel region.
• Various structures have been proposed for imparting a compressive stress or tensile stress to such transistors. In some cases, it has been proposed to provide one or more stressor elements in proximity with an NFET or PFET for applying a beneficial stress to the transistor. For example, commonly assigned U.S. Patent Publication No. 2004/0113174 describes a way of embedding dielectric stressor elements in isolation regions at exterior edges of an active semiconductor region which houses an NFET or a PFET. In such case, the dielectric stressor element and the isolation region are merged. While enabling efficiencies, these isolation-stressor elements require that a design point be reached in which potentially conflicting requirements for the stress-applying function, the isolation function and the processing needed to fabricate them are all simultaneously satisfied.
• Thus, according to the known art, dielectric stressor elements used for applying stresses to an NFET or PFET are constrained to the locations at which isolation regions are placed. To overcome this constraint, it is clear that further improved structures and processing are awaited.
SUMMARY OF THE INVENTION
• The structures and methods provided according to embodiments of the invention herein allow the locations of dielectric stressor elements used with a PFET or an NFET, e.g., the placement, dimensions, edges, etc., of such stressor element, to not be constrained to locations of isolation regions used to isolate the PFET or NFET. Thus, according to one embodiment of the invention, a stress is applied to a channel region of the FET by a “buried” dielectric stressor element. By similar processing, other dielectric stressor elements in accordance with an embodiment of the invention are surface elements provided at a major surface of an active semiconductor region. Preferably, these surface dielectric stressor elements function as combined isolation-stressor elements in locations where isolation regions would ordinarily be provided.
• In accordance with embodiments of the invention, a shear stress is applied to the channel region of the FET by a combination of buried and surface dielectric stressor elements. Thus, in accordance with an embodiment of the invention, on one side of an FET, for example, on a side of the FET on which a source region is disposed, a buried dielectric stressor element extends horizontally under a portion of the active semiconductor region, the dielectric stressor element having an upper surface which underlies the active semiconductor region. An edge of the buried dielectric stressor element that is shared with the active semiconductor region extends in a direction away from the upper surface. According to a preferred embodiment of the invention, such edge can be made closer to the channel region of the PFET or NFET than the edge of a trench isolation region could be located. In addition thereto, a surface dielectric stressor element is provided at a major surface of the active semiconductor region on another side of the FET, that is, on a side preferably opposite the side on which the buried dielectric stressor element is disposed. In combination, the buried dielectric stressor element and the surface dielectric stressor element apply stresses in opposite directions to the channel region of the FET to apply a shear stress thereto.
BRIEF DESCRIPTION OF THE DRAWINGS
• FIG. 1A is a face-up plan view of an FET in accordance with an embodiment of the invention;
• FIG. 1B is a sectional view alongline1B-1B of the FET shown inFIG. 1A;
• FIG. 1C is a sectional view alongline1B-1B of a variation of the FET shown inFIG. 1A;
• FIG. 2 is a face-up plan view of a FET in accordance with another embodiment of the invention.
• FIG. 3A is a face-up plan view of a FET in accordance with yet another embodiment of the invention.
• FIG. 3B is a sectional view alongline3B-3B of the FET shown inFIG. 3A.
• FIG. 3C is a sectional view alongline3B-3B of a variation of the FET shown inFIG. 3A.
• FIGS. 4 through 7 are sectional views illustrating a process of making a FET as illustrated inFIG. 1A, the particular section corresponding toline1B-1B ofFIG. 1A.
DETAILED DESCRIPTION
• New ways of applying a compressive stress and/or a tensile stress to the channel region of a PFET or an NFET transistor are provided according to the embodiments of the present invention that offer simple processing and which are integratable into present methods of manufacturing PFET and NFET transistors of integrated circuits or “chips”. According to the various embodiments of the invention described herein, FETs are provided in various forms in which at least one buried dielectric stressor element underlying one portion of the active semiconductor region exerts a stress upon the channel region of the FET in a first direction and in which at least one surface dielectric stressor element provided at a surface of the active semiconductor region exerts a stress upon the channel region in a second direction opposite to the first direction.
• FIG. 1A is a face-up plan view illustrating aPFET100 in accordance with an embodiment of the invention, in which a surfacedielectric stressor region150 applies a compressive stress in afirst direction156 to a channel region (not visible inFIG. 1A) of the PFET and a burieddielectric stressor region152 applies a compressive stress in a second, opposite, direction to the channel region of the PFET. As illustrated inFIG. 1A, anactive semiconductor region104 of the PFET is bounded partially by anisolation region106, the isolation region being, illustratively, a shallow trench isolation (“STI”) region. The STIregion106 thus defines the boundaries or “edges” of theactive semiconductor region104.
• In describingPFET100, it is helpful to provide a frame of reference in relation to which the elements of the PFET and the buried dielectric stressor elements are located. The directions of a compass: i.e., north, south, east and west provide a useful frame of reference for describing the PFET. These directions are indicated by thelegend101. These directions need not coincide with the true north, south, east and west directions, since thePFET100 can operate in any orientation, at any angle towards the true north direction. Rather, the directions indicated by thelegend101 are useful in describing the placement and orientation of various elements of thePFET100 in relation to each other.
• As defined by STIregion106, the edges of theactive semiconductor region104 include awest edge108 and aneast edge110 which is opposite the west edge in the longitudinal (east-west)direction112 of the PFET. The STIregion106 further defines anorth edge114 and asouth edge116 of theactive semiconductor region104 which is opposite from the north edge in a transverse direction118 of the PFET. As further shown inFIG. 1A, agate120 including agate conductor121 and dielectric sidewalls orspacers123 overlies the active semiconductor region between asource region122 and adrain region124. In the PFET shown inFIG. 1A, the first and seconddielectric stressor elements150,152 apply compressive stresses to theactive semiconductor region104. Specifically, a firstdielectric stressor element150 which is present at the top (major) surface of the active semiconductor region has an inner edge which defines thewest edge108 of theactive semiconductor region104. The second (buried)dielectric stressor element152 is disposed at a predetermined depth from the major surface of the active semiconductor region. The seconddielectric stressor element152 underlies a portion of theactive semiconductor region104 at theeast edge110.
• The compressive dielectric stressor elements shown inFIG. 1A are preferably in form of regions of “expanded oxide,” i.e., regions of oxide which have expanded at least slightly from the volume formerly occupied by semiconductor material of the semiconductor substrate. The first and second dielectric stressor elements apply stresses to the channel region of the PFET in opposite directions, as indicated by thearrows156 and158. The effect of these dielectric stressor elements is to apply oppositely directed stresses to the channel region of the PFET at opposite (west and east) edges108,110 of the active semiconductor region and at different depths from the major surface, such that a shear stress is applied to the channel region.
• FIG. 1B is a sectional view ofPFET100 throughlines1B-1B ofFIG. 1A. As illustrated therein, theactive semiconductor region104 is provided in abulk semiconductor substrate162, preferably being a silicon substrate. A face of theactive semiconductor region104 defines amajor surface105 of the semiconductor substrate. Thegate120, including agate conductor121 andspacers123, overlies achannel region132 of the PFET, as spaced therefrom by agate dielectric125. Edges of thechannel region132 are determined by the locations in the longitudinal direction of afirst edge134 of the gate conductor and asecond gate edge136 opposite thereto. Thesource region122 including an optional extension and/orhalo region126 extends from the vicinity of thefirst gate edge134 to thewest edge108 of the active semiconductor region atSTI region106. Thedrain region124 including an optional extension and/orhalo region127 extends from the vicinity of asecond edge136 of the channel region to theeast edge110 of the active semiconductor region atSTI region106.
• As described above, the firstdielectric stressor element150 has a top surface at themajor surface105 and extends downward therefrom. The first dielectric stressor element has an inner edge which defines the west edge of the active semiconductor region at which asource region122 of the PFET is disposed.
• The seconddielectric stressor element152 has anupper surface140 which is disposed at afirst depth160 from themajor surface105 of the semiconductor substrate. As seen inFIG. 1B, the upper surface of the second (buried) stressor element lies below a thickness of the active semiconductor region, as extends downward from the major surface to the first depth. The second dielectric stressor element underlies a portion of the active semiconductor region in which thedrain region124 is disposed. This contrasts with the first dielectric stressor element which is disposed at the major surface. Specifically, the firstdielectric stressor element150 extends from the major surface to a depth below the major surface, such depth being not substantially greater than the thickness of the active semiconductor region. In order to achieve a desirable stress in a shear direction, the maximum depth of the first dielectric stressor element should not be substantially greater than thefirst depth160. Preferably, the maximum depth of the first dielectric stressor element is the same as or somewhat less than the first depth.
• The second dielectric stressor element does not underlie the entirety of the active semiconductor region, but rather, the first dielectric stressor region shares anedge142 with the active semiconductor region. The edge extends in a direction downward away from the generally horizontalupper surface140. Preferably, theedge142 of the dielectric stressor element is located at about one half a distance between theedge110 of the active semiconductor region and the closest edge (i.e., the second gate edge136) of thegate conductor121,such edge136 being indicated by the dashed line. As discussed above, the effect of the first and second dielectric stressor elements is to apply oppositely directed stresses to the channel region of the PFET at opposite (west and east) edges108,110 of the active semiconductor region and at different depths from the major surface, such that a shear stress is applied to the channel region.
• FIG. 1C depicts a variation of the embodiment described above with respect toFIGS. 1A-1B in which anNFET170 is provided, for which a pair of tensilestressor elements172,174 are disposed at awest edge108 and aneast edge110 of an active semiconductor region. Here,FIG. 1A serves again as a corresponding top-down plan view of the NFET andFIG. 1C is a sectional view ofNFET170 throughlines1B-1B ofFIG. 1A. Unlike the compressive dielectric stressor elements illustrated inFIG. 1B, the tensile dielectric stressor elements shown inFIG. 1C are preferably in form of regions of “collapsed oxide,” i.e., regions of oxide which have shrunken at least slightly from the volume formerly occupied by semiconductor material of the semiconductor substrate. The structures shown inFIG. 1C are the same as those described above with respect toFIGS. 1A-1B, except for the type of transistor (NFET rather than PFET), the dopant types used in each of the respective source regions, drain regions and the channel region in the NFET and the type of stress (tensile) rather than compressive which is applied by eachdielectric stressor element172,174. Thus, as shown inFIG. 1C, a buried tensiledielectric stressor element174, having both atop surface144 and anedge146 in contact with the active semiconductor region, applies a tensile stress in afirst direction186 to the active semiconductor region. On the other hand, a surface tensiledielectric stressor element172 applies a tensile stress in asecond direction184 to the active semiconductor region. The stresses applied by the two dielectric stressor elements combine to apply a shear stress to thechannel region182 of the NFET which has a tendency to “twist” thechannel region182 in a direction as shown byarrow188.
• The PFET depicted inFIGS. 1A, 1B is described above as utilizing compressive stressor elements and the NFET depicted inFIGS. 1A, 1C is described as utilizing tensile stressor elements. However, it is not required that a PFET utilize only compressive stressor elements or that an NFET utilize only tensile stressor elements. In alternative embodiments, it is also possible to provide tensile stressor elements for a PFET at locations corresponding to those described above with respect toFIG. 1C and it is possible to provide compressive stressor elements for an NFET at locations corresponding to those described above with respect toFIG. 1B. While it is to be expected that performance of the individual PFET having tensile stressor elements will not be as great as an individual PFET which has compressive stressor elements, overall performance can still benefit when both PFETs and NFETs of a chip have tensile stressor elements in relation to a chip which lacks such stressor elements.
• Overall performance of a chip can still benefit when both PFETs and NFETs of a chip have compressive stressor elements or when both PFETs and NFETs of a chip have tensile stressor elements. The beneficial effect of the shear stresses applied to the FET may overcome or at least mitigate the effect of somewhat compressive stress being applied to an NFET or somewhat tensile stress being applied to a PFET. Indeed, such arrangement in which PFETs and NFETs both have the same type of stressor elements can be of advantage for some applications, because it requires less processing steps to manufacture than one in which both tensile and compressive-type stressor elements are provided. Particularly in complementary metal oxide semiconductor (“CMOS”) chips, it may be expedient to provide only one type of (tensile or compressive) stressed element in a particular chip. In such case, a net benefit derived from the shear stress applied to either the PFET or NFET may justify the less preferred compressive type of stress being applied to the NFET or the less preferred tensile type of stress being applied to the PFET.
• FIG. 2 is a face-up plan view of aFET200 according to a variation of the embodiment shown and described above with reference toFIGS. 1A and 1B. With respect toFET200, the burieddielectric stressor element250, being disposed at a location similar to and having a construction similar to the buried dielectric stressor element152 (FIG. 1A) underlies thesource region222 of theFET200 at awest edge208 of the FET. In addition, a surfacedielectric stressor element252 has construction similar to and extends from a major surface of the active semiconductor region in a manner similar to the surfacedielectric stressor element150 described above with reference toFIG. 1A. Otherwise, all features of the transistor and the burieddielectric elements250,252 are the same or similar to those ofPFET100 shown and described above (FIGS. 1A and 1B). When theFET200 is a PFET, the dielectric stressor elements preferably have compressive stresses. On the other hand, when theFET200 is an NFET, the dielectric stressor elements preferably have tensile stresses.
• FIG. 3A is a face-up plan view of anFET300 according to another embodiment of the invention. TheFET300 according to this embodiment of the invention is similar to that of PFET100 (FIGS. 1A, 1B) except as to the locations of the buried dielectric stressor element and the surface dielectric stressor element. As depicted inFIG. 3A, a burieddielectric stressor element352 underlies anorth edge314 of the active semiconductor region and portions of thesource region322,drain region324 and channel region (hidden from view as underlying gate conductor321). A surfacedielectric stressor element350 is disposed at the major surface of the active semiconductor region at thesouth edge316 thereof.
• FIG. 3B further depicts a sectional view ofFET360 throughline3B-3B ofFIG. 3A. In the particular embodiment shown inFIG. 3B, theFET360 is an NFET and the dielectric stressor elements are tensile such that they apply tensile stresses intransverse directions356,358 of the FET (direction of a width of the channel382). Again, similar to that shown and described above with respect toFIG. 1C, the buried tensiledielectric stressor element352, having both atop surface344 and anedge346 in contact with the active semiconductor region, applies a tensile stress in afirst direction358 to the active semiconductor region. On the other hand, a surface tensiledielectric stressor element350 applies a tensile stress in asecond direction356 to the active semiconductor region. The stresses applied by the two dielectric stressor elements combine to apply a shear stress to thechannel region382 of the NFET, such that thechannel region382 has a tendency to “twist” in a direction as shown byarrow388.
• FIG. 3C depicts a variation of the embodiment described above with respect toFIG. 3A.FIG. 3C depicts a sectional view of aPFET370 throughline3B-3B ofFIG. 3A. In the particular embodiment shown inFIG. 3C, the surfacedielectric stressor element372 and the burieddielectric stressor element374 are compressive such that they apply compressive stresses in first and secondtransverse directions376,378 of the FET, respectively (direction of a width of the channel392). The stresses applied by the two dielectric stressor elements combine to apply a shear stress to thechannel region392 of the PFET, such that thechannel region392 has a tendency to “twist” in a direction as shown byarrow398.
• Referring now toFIG. 4 throughFIG. 7, a method will now be described for manufacturing theFET100 described above with reference toFIGS. 1A, 1B as an example of manufacturing any of the embodiments of the invention described above. Such method utilizes a process similar to that described in commonly assigned U.S. Patent Publication No. 2005/0067294 to Choe et al. In Choe et al., a region of a silicon substrate is implanted and treated to form a buried oxide layer of a silicon-on-insulator (“SOI”) substrate. A porous silicon region is formed by ion implantation of a p-type dopant (for example, Ga, Al, B and BF₂) and subsequent anodization. The porous silicon region is then oxidized to form the buried oxide layer.
• In the present process, a buried dielectric stressor element is formed at a location of a semiconductor substrate, e.g., a silicon substrate, which underlies only a portion (not all) of an active semiconductor region.FIG. 4 depicts a sectional view corresponding toline1B-1B ofFIG. 1A. As shown inFIG. 4, amasking layer400, e.g., a photoresist is patterned, and a buried region underlying a major surface207 of thesubstrate162 is implanted with a p-type dopant to form a “pocket” p-dopedregion402. As implanted, the dopant concentration inregion402 can range from about 1×10¹⁹cm⁻³to about 5×10²⁰cm⁻³or higher. In any case, the achieved boron concentration must be significantly higher, i.e., one or more orders of magnitude higher than a normal (p-) p-type dopant concentration in the single-crystal silicon. Preferably, the dopant consists essentially of boron (B) or boron fluoride (BF₂), but gallium (Ga) and aluminum (Al) can be used instead. The depth to which ions are implanted into thesemiconductor substrate162 determines the thickness of the dielectric stressor element. The depth of the implant, in turn, is selected in accordance with the energy at which the implant is conducted. As this implant is performed through a photolithographically patterned masking layer, the process of implanting theregion402 defines theedge403 of the implanted region,such edge403 extending in a direction away from the horizontalupper surface401 of the implanted region.
• Thereafter, a surface implantedregion422 is formed at a surface location of the semiconductor substrate, extending downward into thesubstrate162 from themajor surface407, as shown inFIG. 5. As depicted at this stage of the process, asecond masking layer410, e.g., a photoresist, is deposited and patterned, and theregion422 is implanted with a p-type dopant to form a surface p-doped region, using a process similar to that described above with reference toFIG. 4.
• Thereafter, thesecond masking layer410 is stripped and the semiconductor substrate undergoes an anodization process to convert the pocket p-dopedregion402 and the surface p-dopedregion422 into buried porous semiconductor regions. The pocket p-dopedregion402 and the surface p-doped region become porous semiconductor regions as a result of the anodization process.
• The anodization process is as follows. Thesemiconductor substrate162, which preferably consists essentially of silicon and which has buried p-type implanted pocket regions is placed or preferably submerged in a bath containing a solution of hydrogen fluoride (HF), as well as a platinum electrode. Thesemiconductor substrate162 is connected to a positive terminal of a current source, and the platinum electrode is connected to the negative terminal of that current source is connected in conductive communication with the current source that is connected to the positive terminal. The current source supplies an anodization current to the semiconductor substrate and the HF solution which controls the anodization process. In the presence of the anodization current, the HF solution readily diffuses through the single crystal semiconductor (silicon) to the higher concentration p-type doped pocket regions.
• In those higher concentration pocket regions, the HF solution reacts with the highly doped p-type silicon to formporous silicon regions442,444 (FIG. 6) at locations of the implantedregions402,422 shown inFIG. 5. This step is performed prior to forming anadditional masking layer408 as will be described with reference toFIG. 6 below. The anodization current is in the range of 1 mA/cm²to 100 mA/cm², depending on the degree of porosity or density of theporous silicon regions442,444 which are to result from this process. Both the concentration of boron or other p-type dopant in the silicon and the magnitude of the anodization current can be used to control the degree of porosity. That is, these parameters control the density of the resulting porous silicon region, as measured by the mass of silicon that remains within the porous silicon regions divided by their respective volumes. For example, a low porosity region, that is, a region having a relatively high density, is one which has a density of greater than about 44% of the density of the original silicon substrate. On the other hand, a high porosity region, that is, a region having a relatively low density, is one which has a density of less than about 44% of the density of the original silicon substrate.
• After anodization, the substrate is then subjected to a hydrogen bake, which removes most of the implanted boron remaining in the silicon. It is necessary to eliminate high concentrations of boron from the silicon substrate at this stage in order to avoid such high concentrations from interfering with processes used to subsequently define the differently doped regions of a transistor, i.e., the channel region, the source and drain regions, halo and/or extension regions. The hydrogen bake is conducted at temperatures ranging from about 800 degrees centigrade (“C”) to 1,000 degrees C., for periods ranging from about 30 seconds to 30 minutes.
• After the anodization and post-bake processes,regions442,444 (FIG. 6) of porous silicon remain in locations which are at least generally coextensive with the former implantedregions402,422. The porous silicon regions are regions which contain a multiplicity of voids. As viewed with an electron microscope, the porous silicon regions have an appearance similar to a sponge or foam material, having large numbers of voids which are supported together by connecting structure of the remaining silicon material. The degree of porosity in the porous silicon regions is determined at least in part by the initial concentration of boron within the buried pocket regions. As described above, by appropriately selecting the dose of boron that is implanted into the pocket regions and/or by controlling the amount of anodization current, it is possible to remove little mass or much greater mass of the silicon material from the buried pocket regions.
• As further illustrated inFIG. 6, afurther masking layer408, e.g., a hardmask, e.g., silicon nitride, is deposited and patterned over themajor surface407 of the substrate. Thesubstrate162 is then patterned with this masking layer to form atrench415 in an upper region ofsilicon406 above the buriedporous region442 to define anedge110 of the active semiconductor region. Thetrench415 is etched in a location which exposes the buriedporous silicon region442.
• Thereafter, themasking layer408 is stripped and edges of the active semiconductor region are protected appropriately, as by formingspacers412 of silicon nitride thereon, as depicted inFIG. 7. Subsequent thereto, as both the buried porous silicon region and the surface porous region are now exposed from at least top surfaces thereof, both of the exposed porous silicon regions are subjected to an oxidation process which forms thedielectric stressor elements150,152 described above with reference toFIG. 1A.
• Depending on the degree of porosity within the porous regions, the dielectric stressor elements apply a compressive stress or a tensile stress to adjacent portions of the semiconductor substrate. This result is explained as follows. The volume of silicon dioxide is greater than silicon by a ratio of 2.25:1. Thus, when the proportion of silicon that remains within each porous silicon region is greater than 1/2.25 (i.e., the remaining silicon mass within the volume of the porous silicon region is greater than about 44% of the original mass), the resulting silicon dioxide expands, causing the dielectric regions to become compressively stressed when the porous regions are oxidized. Stated another way, the resulting silicon dioxide expands to become compressively stressed when the porosity is less than 56%, that is, when the amount of mass removed from the defined volume of the porous silicon region is less than 56% of the original mass.
• Conversely, when the porosity is greater than 56%, the resulting silicon dioxide contracts, causing the resulting dielectric regions to become tensile stressed. As mentioned above, the degree of porosity is at least partly determined by the conditions under which the regions are implanted with boron and the conditions of the anodization process. In general, the degree of porosity is higher when the implanted boron concentration is higher, and the degree of porosity of lower when the implanted boron concentration is lower. Also, in general, higher porosity can be achieved when the current density of the anodization process is higher. Conversely, lower porosity is achieved when the current density is lower.
• In the processes described in the foregoing, the edges of the implanted regions are defined lithographically. Accordingly, it follows that the extent of the porous silicon regions are determined at least in part by such lithographic processing. Hence, locations of the edges of the dielectric stressor regions that result from oxidizing the porous silicon regions are determined at least in part by the lithographic processing used to mask the substrate when implanting the dopant to form the implanted regions.
• After forming the surface and buried dielectricstressor elements150,152 (FIG. 1A) in the above manner, the trench415 (FIG. 7) is filled with a dielectric material such as an oxide of silicon (e.g., silicon dioxide) to form a trench isolation (“IT”) regions or shallow trench isolation regions (“STI”)region106, as shown inFIG. 1A according to known prior art processing. Such prior art processing typically includes filling the trench with an oxide dielectric, performing chemical mechanical polishing (“CMP”) or an etch back process to reduce the thickness of the deposited oxide to the top of the hardmask (nitride) layer, followed by stripping the remaining nitride hardmask, at the conclusion of which the structure shown inFIG. 7 results. In such process, the dielectric fill is deposited via a high density plasma (“HDP”) technique and/or other chemical vapor deposition (“CVD”) technique including low pressure CVD (“LPCVD”), plasma enhanced CVD (“PECVD”), etc., which may involve deposition form a tetraethylorthosilicate (“TEOS”) precursor, for example. The dielectric material can include a nitride, e.g., silicon nitride which lines interior walls of the trenches, prior to deposition of the dielectric fill.
• After forming the buried dielectric stressor elements, thegate conductor121,dielectric spacers123, and source and drainregions122,124, including extension regions and/orhalo regions126,127 are formed, as shown inFIGS. 1A-1B. This completes the formation ofFET100, having thedielectric stressor elements150,152 as shown in the sectional view inFIG. 1A.
• While the invention has been described in accordance with certain preferred embodiments thereof, those skilled in the art will understand the many modifications and enhancements which can be made thereto without departing from the true scope and spirit of the invention, which is limited only by the claims appended below.

Claims (15)

1. A chip, comprising:

an active semiconductor region having a major surface and a thickness extending from said major surface to a first depth below said major surface;

a field effect transistor (“FET”) having a channel region, a source region and a drain region all disposed within said active semiconductor region, a length of said channel region being oriented in a longitudinal direction of said active semiconductor region, and a width of said channel region being oriented in a transverse direction of said active semiconductor region transverse to said longitudinal direction;

a first dielectric stressor element laterally adjacent to a first edge of said active semiconductor region, said first dielectric stressor element extending from said major surface of said active semiconductor region downward to a depth not substantially greater than said first depth; and

a second dielectric stressor element underlying only a portion of said active semiconductor region at a second edge of said active semiconductor region opposite said first edge, said second dielectric stressor element having a horizontally extending upper surface at said first depth, said second dielectric stressor element sharing an edge with said active semiconductor region, said edge extending in a direction away from said upper surface,

said first dielectric stressor element applying a first stress to said channel region in a first direction and said second dielectric stressor element applying a second stress to said channel region in a second direction opposite to said first direction such that said first and second stresses cooperate together to apply a shear stress to said channel region.

2. The chip as claimed inclaim 1, wherein said first dielectric stressor element exerts a compressive stress in said first direction, and said second dielectric stressor element exerts a compressive stress in said second direction.

3. The chip as claimed inclaim 1, wherein said first dielectric stressor element exerts a tensile stress in said first direction, and said second dielectric stressor element exerts a tensile stress in said second direction.

4. The chip as claimed inclaim 1, wherein said active semiconductor region has a west edge and an east edge remote from said west edge in a longitudinal direction of said active semiconductor region, and said active semiconductor region has a north edge and a south edge remote said north edge in a transverse direction of said active semiconductor region, said second dielectric stressor element contacting a trench isolation region, said trench isolation region sharing at least one of said north edge, said east edge, said south edge and said west edge with said active semiconductor region.

5. The chip as claimed inclaim 1, wherein said edge of said second dielectric stressor element extends in a direction away from said upper surface of said second dielectric stressor element.

6. The chip as claimed inclaim 4, wherein said FET further includes a gate conductor having a conductive portion overlying said channel region, said conductive portion having a first vertically oriented gate edge and a second vertically oriented gate edge opposite said first gate edge, wherein said edge of said second dielectric stressor element is disposed about half a distance between an edge of said trench isolation region and said second gate edge.

7. The chip as claimed inclaim 4, wherein each of said north, south and east edges of said active semiconductor region are shared with said trench isolation region, said second dielectric stressor element contacts substantially less than entire lengths of said north edge and said south edge and said first dielectric stressor element functions to isolate said west edge of said active semiconductor region.

8. A chip, comprising:

an active semiconductor region having a west edge, an east edge, a north edge and a south edge, said active semiconductor region having a longitudinal direction in a direction between said west and east edges and a transverse direction in a direction between said north and south edges, said active semiconductor region having a major surface and a thickness extending from said major surface to a first depth below said major surface,

a field effect transistor (“FET”) having a channel region, a source region and a drain region all disposed within said active semiconductor region, a length of said channel region being disposed in said longitudinal direction, and a width of said channel region being disposed in said transverse direction;

a first dielectric stressor element disposed laterally adjacent to a first edge of said active semiconductor region, said first edge including at least one of said north edge, south edge, east edge or west edge, said first dielectric stressor element extending downward from said major surface of said active semiconductor region to a depth not substantially greater than said first depth; and

a second dielectric stressor element underlying a second edge including at least one of said north, south, east or west edges of said active semiconductor region, said second edge being remote from said first edge, said second dielectric stressor element having a horizontally extending upper surface at said first depth, said second dielectric stressor element sharing a third edge with said active semiconductor region, said third edge extending in a direction away from said upper surface,

said first dielectric stressor element applying a first stress to said channel region in a first direction and said second dielectric stressor element applying a second stress to said channel region in a second direction opposite to said first direction, such that said first and second stresses cooperate together to apply a shear stress to said channel region.

9. The chip as claimed inclaim 8, wherein said first dielectric stressor element exerts a tensile stress in said first direction, and said second dielectric stressor element exerts a tensile stress in said second direction.

10. The chip as claimed inclaim 9, wherein said first dielectric stressor element exerts a compressive stress in said first direction, and said second dielectric stressor element exerts a compressive stress in said second direction.

11. A method of fabricating a field effect transistor (“FET”) device, comprising:

forming a horizontally extending buried porous semiconductor region having an upper surface at a first depth below a major surface of a portion of a horizontally extending active semiconductor region of a substrate, said buried porous semiconductor region and said active semiconductor region including a first semiconductor, said buried porous semiconductor region having a multiplicity of voids and having a first density selected by a parameter of a process of forming said buried porous semiconductor region, said first density being substantially lower than a second density of said active semiconductor region;

forming a surface porous semiconductor region extending from said major surface to a second depth not substantially greater than said first depth on a side of said active semiconductor region opposite said buried porous semiconductor region, said surface porous semiconductor region including said first semiconductor, said surface porous semiconductor region having a multiplicity of voids and having said first density;

oxidizing said first semiconductor included in said buried porous semiconductor region and in said surface porous semiconductor region to form buried and surface dielectric stressor elements, respectively;

forming a field effect transistor (“FET”) having a channel region, a source region and a drain region all disposed within said active semiconductor region,

wherein said buried and surface dielectric stressor elements apply either compressive stresses or tensile stresses upon said channel region of said FET, said first density of said porous semiconductor region determining whether said stress is compressive or tensile, said stresses applied by said buried and surface dielectric stressor elements cooperating together to apply a shear stress to said channel region of said FET.

12. The method of manufacturing the semiconductor device as claimed inclaim 11, wherein said steps of forming said buried and surface porous semiconductor regions include implanting a dopant into regions of said substrate through an opening in a mask and exposing said substrate to an anodization process.

13. The method of manufacturing the semiconductor device as claimed inclaim 12, wherein said steps of forming said buried and surface porous semiconductor regions further include a pre-bake process to reduce a concentration of said dopant within said buried and surface porous semiconductor regions.

14. The method of manufacturing the semiconductor device as claimed inclaim 12, wherein said region of said semiconductor substrate into which said dopant is implanted in forming said buried porous semiconductor region underlies said active semiconductor region when said step of implanting is performed.

15. The method of manufacturing the semiconductor device as claimed inclaim 12, wherein said edge of said buried dielectric stressor element is determined photolithographically during said step of implanting said regions of said semiconductor substrate.

Images